Comparing Q1 FY2026 (April 14, 2026) → Q2 FY2026 (July 14, 2026)
Delivered
Consumer Resilience: The Forecast Held
In April, Jeremy Barnum’s assessment of the American consumer was notable mainly for its absence of new information. Nothing had cracked. Every angle they looked at—early roll rates, delinquency rates, cash buffers, discretionary spend—showed, as he put it, “fundamentally healthy” behavior. The caveat was equally explicit: the labor market was doing the work, and if that changed, everything else would follow.
Three months later, the language held almost word for word. Consumers and small businesses “continue to show resilience.” Spend growth remained strong. Delinquencies came in “a little lower than we expected.” The card net charge-off rate was revised down from 3.4% to 3.2%, a direction that required no hedging. Barnum noted that the performance looked “pretty much across the board by FICO score,” which is a harder thing to say than a single-cohort improvement.
The delivery here was clean. The April commitment—consumer health contingent on labor—arrived in July with the labor market described as “surprisingly resilient.” The conditionality was preserved rather than erased, which is a form of intellectual honesty. Management did not claim the outcome was guaranteed; they simply reported that the condition held and the forecast followed.
The bet on the labor market paid, at least for now.
NII Guidance: Revised Up, Both Buckets
April’s NII guidance held two numbers: NII ex-markets of approximately $95 billion for the full year, and total NII of approximately $103 billion. The rationale for not revising up despite higher rates was laid out with precision—the EAR of $1.8 billion, a roughly 20 basis point impact on the full-year average from the backdated cuts, rounding effects. Barnum’s conclusion: “I don’t think there’s too much to read into it.”
In July, both numbers moved higher. NII ex-markets revised to $96.5 billion; total NII to $105.5 billion, with markets NII now projected at approximately $9 billion. Barnum identified the primary driver as deposit balances—”the biggest single factor”—across both wholesale and consumer, including mix shift toward slightly higher margins. Rates contributed, but less than the balance story.
One nuance worth noting: markets NII moved up despite rates being higher—which should have pushed it down, given the liability-sensitive exposure. The explanation was balance sheet composition, specifically a reduction in non-interest-bearing assets in the second half. Barnum also noted a small mechanical contribution from the equity reallocation to CIB, roughly $150 million that would not offset in NIR. This kind of precision is consistent with the April tone. Nothing was dressed up.
Investment Banking: Pipeline Resilience Confirmed
The April characterization of the IB pipeline was measured: “resilient, maybe surprisingly resilient.” Barnum had flagged that some of the first-quarter outperformance reflected accelerated M&A closure from faster-than-expected regulatory approval—a timing pull-forward, not a trend signal. The pipeline going forward was described as healthy but conditional on geopolitical outcomes.
In the second quarter, IB fees were up 30% year-on-year, with double-digit growth across all products and particularly strong equity underwriting. The pipeline description moved from “resilient” to “quite robust.” Barnum was transparent about the second-quarter structure: there was some pull-forward, some large ECM deals contributed meaningfully, and some exceptional events inflated the results. But his read on forward momentum was that the high-profile activity was itself “begetting more activity”—deal confidence feeding deal confidence.
The April commitment was to pipeline resilience; July delivered both confirmation and something more. Whether the word “quite” in “quite robust” carries weight is the open question here. It is one word stronger than what was said three months ago.
Reprioritized
Excess Capital: From Patience to Motion
The April framing of excess capital was a study in deliberate restraint. The number was approximately $40 billion. Deployment preference was organic—more bankers, global banking, opening countries. Buybacks were acknowledged but framed as second-best: “We prefer to deploy the capital serving clients.” Dimon specifically said he wanted to remove the “cash returned to investors” metric from reporting because it suggested buybacks were inherently good, when they weren’t. The posture was: we have it, we don’t feel rushed, we’d rather use it for business.
By the July call, that same $40 billion figure reappeared—but the tone around it had shifted. Dimon said, “I think we now think we actually deployed it over time. The world’s gotten bigger.” He mentioned a recent European tour, the hyperscaler investment needs, global infrastructure and remilitarization. The emphasis was on organic opportunity scale rather than capital return mechanics. He also mentioned a prior comment about $20 billion in potential activity—calling it a mistake to have named a number—and declined to re-specify, saying “we could obviously do far more than that or nothing at all.”
What changed is the frame. In April the question was what to do with the capital. In July the implication was that the capital is already finding its uses, even if management won’t specify them. The $40 billion didn’t change; the framing of whether it needs a home changed.
🟠 Narrative Inertia: The language around excess capital is consistent in quantity but drifting in urgency. April was patient; July was active without specifying action. The dollar figure hasn’t moved; the implied pressure on it has.
Regulatory Capital: Continued Advocacy, Adjusted Tone
The April prepared remarks gave substantial airtime to the Basel III Endgame and G-SIB reproposals—arguably the most specific policy argument made in the presentation. Barnum laid out the arithmetic: JPMorgan’s position was worse than the Fed’s aggregate estimate for Category 1 and 2 banks in each subcategory. The G-SIB surcharge trajectory pointed to 5.2% by 2028, a 70 basis point increase, implying roughly $20 billion of additional capital requirements. The short-term wholesale funding methodology was called out as adding $13 billion specifically to JPMorgan while reducing risk-sensitivity. The phrase “persistent miscalibration” was used.
The July treatment of the same topic was sharper and shorter. Dimon named four specific changes: do the numbers correctly without artificial inflation, eliminate the double-count in operating risk capital, eliminate the double-count in market risk capital, and fix the G-SIB short-term wholesale funding methodology to match the original 2015 design. He also made the framing explicit: if regulators want more capital, they should simply say so directly. “I’m not happy to have these numbers falsely done.”
The policy position is unchanged. What moved is the forum. April was investor-facing disclosure of the problem. July was a cleaner, harder version of a demand for resolution—naming the four specific fixes rather than the multi-slide architecture of the earlier call. Barnum added a pointed observation about competitive dynamics: the short-term wholesale funding change disproportionately burdens institutions with both markets and consumer businesses relative to pure-play investment banks. Someone could want that outcome, he said. But they should say so.
Smart Cash Tool: From Experiment to Commitment
April framing from Barnum: “I think the right way to think of it is sort of as an experiment right now.” The product wasn’t live yet. It was described as targeted at a narrow subset of clients with both investments and deposit relationships. Dimon’s read was similarly low-key: it’s about giving people better tools to manage what they already have across accounts, and competition for deposits has always been intense.
The July exchange showed movement without dramatic escalation. Dimon confirmed the tool was still in test phase—”certain tests coming out”—and explicitly said “you’ll see something this year.” The language remained cautious about scope: “these don’t relate to every account.” But the commitment frame shifted. April was “an experiment.” July was “we think it could be good for customers and good for us” and a confirmed 2026 timeline.
Whether this constitutes material progress or simply maintained intention is genuinely unclear from the transcript alone. The tool remains narrow in scope. The timeline was already implied in April. What changed is the commitment language: “you’ll see something this year” is a harder statement than “it’s sort of an experiment.”
De-Emphasized or Absent
The Middle East Risk: Gone
In April, geopolitical risk from the Middle East conflict was woven into nearly every forward-looking statement. Client engagement in investment banking was “healthy, but of course, developments in the Middle East could have an impact on deal execution and timing.” Consumer resilience was qualified by the possibility of higher energy prices from the conflict filtering into the labor market. Reserve methodology discussion explicitly included “higher energy prices that wind up having an impact on the core global economic outlook” as a downside scenario being monitored. Barnum described IB pipeline sentiment as resilient despite “everything that’s going on,” and specifically noted that timelines in the region created some ambiguity about deal decision-making.
The July call contained no equivalent language.
The geopolitical risk category that occupied meaningful space in the first quarter was absent from prepared remarks and did not arise in Q&A in any materially comparable form. The May-through-June period involved meaningful developments in the region, but the July call did not qualify forward statements with references to deal timing, energy price transmission, or reserve scenario weighting in the same way. What was present in April as a live variable appears to have been set aside, at least for earnings call purposes.
🔴 Evaporated Narrative: The Middle East conflict qualifier was present and specific in April—embedded in consumer, reserve, and IB pipeline discussions. It is absent from the July call without acknowledgment. Management may have concluded the risk pathway did not materialize; alternatively, there may be a decision to not re-flag it absent new negative developments. The absence itself carries information. (Moderate Confidence)
Private Credit Systemic Risk: Compressed to a Single Comment
April gave private credit substantial attention. Dimon addressed the topic across multiple Q&A exchanges: sizing the exposure ($60 billion in what he termed “back leverage”), discussing whether defaults in the $1.7 trillion market could be systemic (no, but credit cycle losses would be worse than expected), describing the underwriting discipline that protected JPMorgan’s position, and explaining the structural protections embedded in the exposure. Barnum walked through the NBFI exposure walk from $330 billion to $160 billion to the $50-60 billion core private credit number. The topic generated multiple follow-up questions and extended answers.
In July, private credit appeared in one short Dimon comment on credit underwriting standards—describing modest deterioration (”more PIK,” “some weaker covenants,” “a little bit of weakness across that spectrum”) and reiterating the credit cycle thesis that losses would exceed expectations and performance would not follow a bell curve. There was no sizing, no structural breakdown, no NBFI framework discussion.
This is reduction in emphasis, not disappearance. The credit cycle caution was restated. But the scaffolding built around it in the first quarter—numbers, structure, specific protections, systemic vs. non-systemic analysis—was not reconstructed. The April call had something to prove on this topic. The July call did not re-argue it.
Reserve Methodology: The Conscious Debate, Quietly Resolved
April contained an unusually candid disclosure about the allowance-building process. Barnum described a “very conscious debate” the company had internally about whether to add downside skew to scenario weights given the geopolitical environment. The conclusion was no—the existing conservative bias was sufficient, and they would wait for actual deterioration to flow through the model. He named the specific mechanism: an unemployment forecast revision from 5.8% to 5.6% on unchanged weights created some consumer tailwinds.
The July call contained a reserve build of $149 million on $2.4 billion in charge-offs. There was no equivalent discussion of scenario weighting deliberation, no reference to the downside skew debate, no explanation of whether weights changed. The reserve topic arose only incidentally.
It is possible the lack of discussion reflects an unchanged process—the debate was resolved, conditions did not change materially, so there is nothing new to report. It is equally possible that the reserve methodology discussion only generated attention in April because conditions at that moment made it feel important. Whether the deliberate transparency was a one-time event or a commitment to a disclosure pattern is not yet answerable from two data points.
Narrative Positioning
“Getting Close to As Good as It Gets”
The April call framed market conditions through the lens of volatility—specifically that the first quarter had been good volatility, the kind that generates client activity and wider spreads rather than the gappy, illiquid kind that keeps clients sidelined. Dimon’s description of the trading business in April was expansive: “very good business,” “fabulous people,” revenue strength explained through capital deployment and scale. The tone was strong but not triumphalist.
July introduced language that April did not contain. When asked whether the environment was “as good as it gets,” Dimon said: “It’s getting close to as good as it gets. We just don’t know how long it’s going to last.” This is a new register. It is not a warning, but it is a ceiling acknowledgment. In the same call, equities revenue came in up 86% year-on-year—a number Barnum described as unlikely to repeat based on its specific drivers. The combination is deliberate: exceptional results plus a statement that conditions may be near peak.
The intent appears to be expectation management through performance credibility. It is easier to say “this may be the ceiling” when you are reporting the highest equities quarter in recent memory than to say it when results are merely good. The framing takes advantage of the moment.
Succession: New Structure, Unchanged Timeline
Succession was not a named topic in April’s prepared remarks and did not arise in Q&A. It was background context—present in the organization but not under active narrative management on the earnings call.
July opened with the first Q&A question on the co-president announcement and Marianne Lake’s departure. Dimon’s handling was brief and direct: the board made a decision, Lake chose retirement over the new structure, no mystery. When pressed on timetable, his response was to reference what he had said last time—”several years, plus or minus”—and to note that the timetable was unchanged. Erika Najarian pushed for clarification on what “no change” meant, and Dimon restated: “Exactly what we said last time.”
The framing strategy was minimal elaboration. Dimon answered the factual question (timing unchanged, two co-presidents now prepared for broader roles) and declined to add color on the internal dynamics beyond “no mystery.” The repetition of “no change” in response to a follow-up question is itself a framing choice—it positions the event as a structural update rather than a signal.
Q&A: Risk Calibration Under Pressure
Ken Usdin asked, after reviewing the quarter’s results, how risk-on JPMorgan itself was being in this environment. Barnum’s answer was careful and revealing:
“Not to be pedantic, but I think the question, the ‘we’ matters, right? The market is clearly extremely risk on, and we’re kind of takers of that. We’re trying to strike the right balance between supporting all our clients and being appropriately cautious in an environment that has some complicated dynamics in it.”
This is a direct reframe. The distinction between the market being risk-on and the bank being risk-on is meaningful. Barnum positioned JPMorgan as a market-neutral intermediary responding to client demand rather than a participant expressing directional risk appetite. “Takers of that” is a specific phrase—it implies passivity relative to the environment rather than active risk accumulation.
April did not contain a comparable formulation because the question wasn’t posed in the same way. In April, the question was whether trading strength was sustainable; Barnum described it as capital deployment with healthy but below-average returns. In July, the question was whether the bank itself was positioned aggressively. The July answer closes that question down without ceding the ground. Whether it fully addressed what the analyst was asking about internal risk standards is a different matter.
AI Narrative: From Adjacent Opportunity to Operational Center
April’s AI discussion was split between risk and opportunity. On risk: cyber was called the largest threat to the firm, with AI specifically named as making it harder. On opportunity: Dimon pushed back on the efficiency ratio narrative—competitive markets would pass AI benefits to customers, not to return on equity—while acknowledging real adjacency value in consumer products, fraud reduction, and prospecting. The AI cash tool was positioned as a customer-facing experiment.
The July AI conversation was bigger, faster, and more internal-facing. Dimon described ~1,000 use cases with the most important 50 across risk, fraud, marketing, and document reading. Specific areas where headcount had been reduced by 30-40% were mentioned as outcomes already delivered, with “most of those people offered jobs elsewhere.” Barnum raised a new topic that hadn’t appeared in April: token expense. He described it as trivial now but with “meaningful acceleration” projected for the second half of the year, and flagged it as an important question for 2027 and beyond. The firm’s AI infrastructure spend was positioned as creating future model-selection sophistication—using the right model for the right task rather than defaulting to expensive frontier models.
Cyber risk, by contrast, was not raised in July prepared remarks and did not arise with comparable prominence in Q&A. In April, Dimon called it JPMorgan’s largest risk and addressed it at length. The specific mention of Anthropic’s Mythos model creating new attack surface in April was notable. By July, the AI narrative had rotated toward deployment and operational efficiency rather than threat management.
🟡 Convenience Pivot (Weak Signal): It is possible the rotation from AI-as-cyber-risk to AI-as-efficiency-driver is simply a function of what questions were asked rather than a deliberate frame shift. Cyber risk was discussed in the context of a direct question in April. No equivalent question dominated the July session. The absence of cyber AI risk discussion in July is more likely a Q&A artifact than a narrative decision. (Low Confidence)
Deposit Competition: Market Share Aspiration, Unchanged
April’s deposit discussion covered the near-term mechanics—yield-seeking flows, the role of tax refunds, 450,000 net new checking accounts, consumer deposit growth expectations of “low to mid-single digits.” Barnum was cautious about calling a trend given the tax season effect.
Three months on, consumer deposits were up 3% year-on-year with over 500,000 net new checking accounts—stronger on both dimensions. The longer-term 15% retail market share aspiration was reaffirmed without revision: “we still feel good about it,” positioned as “a natural long-term consequence of executing the strategy.” Wholesale outperformed the earlier expectation of modest growth.
The 15% target didn’t move, nor did the framing that it is an outcome rather than an objective. The checking account momentum improved; the strategic confidence held. A modest positive trend on the mechanics, no change on the ambition.
One question that neither call fully answered: what does the expense guidance revision actually signal about the second half? Barnum decomposed the $2.5 billion increase clearly—$1.5 billion already booked from first-half capital markets outperformance, $1 billion implicitly added for the back half. But he explicitly said not to draw too many conclusions from the $1 billion about revenue expectations. The guide includes it; the language disclaims it. Analysts know volume-related expenses follow revenue. What the $1 billion in implicitly added second-half expense guidance implies about Barnum’s private revenue forecast for that period is the number that didn’t get named.
The Gap Report is narrative intelligence, not investment advice.
Quotes are verbatim from publicly available earnings call transcripts. This analysis reflects an interpretation of language and tone shifts between calls
Comments
Log in or sign up to join the conversation.